CPU release
tionichm opened this issue · comments
At this point the base torch packages are enormous, and it is affecting the startup time of inference containers in our solution. We've been looking at using torch+cpu releases to work around this, but we have a dependency on this package.
I propose that this package gets split into a main (CPU) release with optional GPU compatibility, or the other way around, where there is a main GPU release and a separate CPU release.
This issue is stale because it has been open 60 days with no activity. Remove stale label or comment or this will be closed in 7 days.
This issue was closed because it has been stalled for 7 days with no activity.